UPDATE: This project won a major bursary prize in the Computer Programs category!
This app was developed for the Science Talent Search Victoria 2021.
Cloudvision is an app designed for learning, collecting, scanning and analysing clouds. Features include:
- A cloud collection gallery for ease of access to saved clouds.
- A cloud scanner that makes use of the device's camera and machine learning to output the cloud type.
- A collection of learning resources about clouds and the creation of the application.
Installation: Install via App Store (iOS): https://apps.apple.com/au/app/cloudvision/id1597605747
Technical info:
Cloudvision was built using the Flutter framework. All front-end code is written in Dart. Cloudvision also makes use of TensorFlow machine learning and the Teachable Machine by Google. A huge thanks to Havard Dataverse's CCSN Database of cloud images that were used for training the machine (https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/CADDPD).
Specs:
- Built on an iMac (27-inch, Late 2012)
- Developed in IntelliJ IDEA CE (IDE) / Xcode (Deployment to iOS)
- Application tested and deployed on an iPhone Xs (iOS 14.6).